Cancel
Showing results for 
Search instead for 
Did you mean: 

Why Rich client performance drops with latency?

Pioneer
Pioneer

Today I had a user asking me to explain in simple terms why does the rich client performance drops in a 4-Tier across WAN. If i have the Application and Web Servers in Site A, then why is 4T Rich Client afster in SiteA than for a user in SiteB who is few thousand Kms apart. I know the answer is latency, but he was probing further as to how latency is causing the slowness when we use the same thin client link. My answer was that the response recieved from server was faster in LAN Vs WAN...but it made me re-think to get a much clearer answer to convince him. Any one has better answers to this?

Also why is a thin client faster than thick (rich) client if done from Site B?

 

The question is very naive - but i prefer to get my thoughts clarified and clear.

2 REPLIES

Re: Why Rich client performance drops with latency?

Solution Partner Honored Contributor Solution Partner Honored Contributor
Solution Partner Honored Contributor
The webclient (thin client) is actually running on the server where pool manager is running and only displaying back some web information. It doesn't handle all situations very well otherwise everyone would be using the thin client (and why they invented Active Workspace which is HTML5).

The Rich Client does some processing locally on the workstation and some processing on the server which gives an advantage with some applications. It is the strongest of all the architectures so far but Active Workspace is a close second and in many instances much faster. Since some of the processing is local then the client needs to ask the server for (meta) data. This query process goes back and forth many many times just to show some simple data. It has to chack access for all the objects, relationships, display rules, etc. These trips take time and the longer the distance between the client and the server then the longer each response takes. Add up the response times over many queries and you begin to see that the further you are away then the worse the performance. Its not just distance but also throughput. For instance, adding WAN accelerators, like Cisco WAS or Riverbed, will only send "deltas" for the meta data which means less traffic and smaller packages to transmit and receive therefore faster performance. Teamcenter used to be VERY talkative and Siemens has done a terrific job of reducing the number of calls and how much data is required for each call over the years. Today it is much more latency tolerant (think back to the 2t rich client days). Active Workspace is the new normal and out performs the 4t rich client on a regular basis. We used to only have the 2tRAC, then they added the thin client, then the 4tRAC and now Active Workspace. Each is better than the previous. Hope that helps.

Randy Ellsworth, Teamcenter Architect, Applied CAx, LLC
NX 11 | SW 2016 | Creo 4 | TcUA 11.4
Evaluating: AW 3.4

Re: Why Rich client performance drops with latency?

Pioneer
Pioneer

Thanks Randy. Nicely explained.